LESSON 6: STATISTICAL TOOLS Flashcards
To know if the research instruments we use are precisely measuring the variables, we investigate
Exploring Instrument Reliability
Measure the internal consistency of the questionnaire.
Cronbach alpha and McDonald’s Omega
Accepted measurement of reliability
0.70 - 0.80
STATISTICAL TOOL TO EXPLORE ASSUMPTIONS
- Check the distribution of data, p>0.05 - normal data, p<0.05 - non-normal data
- Continuous Data
Shapiro Wilk Test
Kolmogorov Smirnov Test
STATISTICAL TOOL TO EXPLORE ASSUMPTIONS
- Assume that variances of the population from which different samples are drawn are equal, p>0.05 - equal variances, p<0.05 - do not assume equal variances.
- Continuous Data
Levene’s Test
STATISTICAL TOOL USED TO EXPLORE RELATIONSHIP
- Tests for the strength of the association between two continuous variables.
- Two Continuous Data
- Normal Data
Pearson Correlation
STATISTICAL TOOL USED TO EXPLORE RELATIONSHIP
- Allows to control for possible effects of another confounding variable (e.g. socially desirable responding).
- Two Continuous Data
- Normal Data
Partial Correlation
STATISTICAL TOOL USED TO EXPLORE RELATIONSHIP
- Tests for the strength of the association between two ordinal variables.
- Ordinal Variable and One Continuous Data
- Non-Normal Data
Spearman Correlation
STATISTICAL TOOL USED TO EXPLORE RELATIONSHIP
- Two Continuous Data
- One Nominal Data
- Normal Data
- Independence from errors
Multiple Regression
STATISTICAL TOOL USED TO EXPLORE RELATIONSHIP
Test the predictive validity of one (1) independent variable on one continuous measure (dv).
One Continuous Data or Nominal Data
Do not assume normality
Ordinary Least Square
Or
Simple Regression
STATISTICAL TOOL USED TO EXPLORE RELATIONSHIP
- Allows to condense a large set of variables or scale items down to smaller, more manageable number of dimensions.
- Continuous Data
- Normal Data
Factor Analysis
STATISTICAL TOOL USED TO EXPLORE RELATIONSHIP
- Tests for the strength of the association between two categorical variables.
- Categorical Data
- Large Sample Size
- Non-normal Data
Chi Square Test for Relatedness/Independence
STATISTICAL TEST USED TO EXPLORE DIFFERENCES
- Tests the difference between two sets of observations (before and after)
- Two Continuous data
- Normal data
T-test for paired sample known as repeated measure
STATISTICAL TEST USED TO EXPLORE DIFFERENCES
- Test the difference between the means of two unrelated independent groups
- Two Continuous data
- Categorical and continuous data
- Normal data
T-test for independent sample
STATISTICAL TEST USED TO EXPLORE DIFFERENCES
- Test the difference between a sample mean
- Ordinal variable and continuous
- Two ordinal data
- Nominal and ordinal variable
- Normal data
T-test for one sample
STATISTICAL TEST USED TO EXPLORE DIFFERENCES
- Test the differences among two or more groups and you wish to compare their mean scores on continuous variable
- Categorical and continuous data
One way analysis of variance
STATISTICAL TEST USED TO EXPLORE DIFFERENCES
- Test the differences among two independent variables. Good for testing interaction effects and main effects
- Two categorical data and one continuous data
Two way analysis of variance
STATISTICAL TEST USED TO EXPLORE DIFFERENCES
- Test for comparing group with groups of number of different but related variables
- Categorical and continuous data
Multivariate analysis of variance
TEST USED FOR NON-NORMALLY DISTRIBUTED DATA
- Test differences on a dichotomous dependent variable between two related groups (before and after)
- Nominal data and Dichotomous data
McNemar Test
TEST USED FOR NON-NORMALLY DISTRIBUTED DATA
- Alternative to Independent Sample T-Test
- Categorical and Continuous data
Wilcoxon rank-sum test
TEST USED FOR NON-NORMALLY DISTRIBUTED DATA
- Alternative to Independent Sample T-Test
- Ordinal data
- Categorical and Continuous data
Mann Whitney U-test
TEST USED FOR NON-NORMALLY DISTRIBUTED DATA
- Alternative to paired sample t-test
- Categorical and Continuous data
Wilcoxon sign-rank test
TEST USED FOR NON-NORMALLY DISTRIBUTED DATA
- Alternative to ANOVA
- Categorical and Continuous data
Kruskal-Wallis Test and Friedman Test